Google has released its annual search spam report, which shows that SpamBrain, a machine learning-based spam identification system, treated six times as many spam websites in 2021 as it did in 2020.
By identifying these sites, Google claims a 70 percent reduction in hacked spam and a 75 percent reduction in gibberish spam on hosted platforms.
By identifying sites with disruptive or malicious behavior, Google has kept 99% of searches spam-free.
SpamBrain demonstrates continuous improvement of Google AI
Although Google unveiled its spam detection system in 2018, this report marks the first external reference to it as “SpamBrain”.
Using machine learning, this AI platform has become more effective at detecting and remediating spam and is evolving to combat new types of spam and prevent abuse.
This includes the ability to identify websites that have been hacked or otherwise compromised in order to install malicious recording software on users’ computers.
Another area of improvement was the identification of gibberish spam sites consisting of meaningless text with keywords.
Black hat SEO experts use these to improve page ranks and sometimes even spend money on spam links to trick search engines into increasing PageRank quality scores.
Google is committed to search quality and user safety
Aimed at improving search quality and protecting user safety, Google claims to be making significant strides in fighting link spam, fraud and online harassment.
A link spam update mid-year improved the search engine’s link rating and improved its ability to detect unnatural links in multiple languages and exclude them from search results.
Last year, consumers lost more than $5.8 billion to fraud, a 70% increase from 2020. To counteract this, Google expanded its user safety program.
Building on the work of the previous year, it achieved a 40 percent reduction in fraudulent search results.
As part of its commitment to user safety, it also changed its approach to online harassment and the way it handles results for requests that are vulnerable to attackers. This includes websites that use pay-to-remove content.
Combating ranking manipulation
Noting the unethical practices some sites employ to circumvent their quality guidelines, Google has also taken steps to reduce ranking manipulation. This includes changing how product reviews are handled and allowing users to report manipulative practices.
In 2021, Google rolled out two major updates to how product reviews are rated:
Google says these updates have been successful in reducing low-quality reviews while encouraging reviews with better content and authorship.
source: Google Search Headquarters
Featured image: smx12/Shutterstock